Dynare forums

I performed a DSGE estimation with 1000,000 posterior draws.
However, one strange thing is that posterior standard deviations of estimated parameters are exactly same as those of prior.

Q: Is there any way I can check simulated series which will be length of 500,000 (because I burned in 50%).

If you are referring to the table column "pstdev" that actually is the prior standard deviation, not the posterior one. For the posterior we only report the 90% HPD interval.

If that is not the problem, check identification.

Thanks for your reply.

Then, is there any way I can compute posterior standard deviations on my own?
In other words, if simulated series are stored, we could calculate them.

Take the square root of

oo_.posterior_variance